Reply to thread

Message

<blockquote data-quote="turnip_farmer" data-source="post: 8189735" data-attributes="member: 7029365">The best way of testing would just be to roll them a whole bunch of times and then do a statistical test on the plausibility of your results.Might take a while if you have a lot of dice. What I would do (and might do this weekend now I'm thinking about it - that's how you kill time in quarantine), is do it with a relatively low number of rolls for each, and then look in more detail at those which look suspicious.A chi-squared test is easy to do. Say you have a four-sided dice and roll out 40 times. You would expect each number to show up approximately 10 times.Count how many times each number appears. For each, subtract 10 (the expected value), square the result, and then divide that by ten. Add your four numbers together, and then find yourself a chi-squared table online (like <a href="https://people.richland.edu/james/lecture/m170/tbl-chi.html" target="_blank">here)</a>You would have 3 degrees of freedom (one less than the dice number), so find 3 on the leftmost column, then read across until you got whatever number you calculated. Looking at the top row at that point will tell you the probability of getting a result at least that far away from the expected value of your dice was fair.</blockquote>

[QUOTE="turnip_farmer, post: 8189735, member: 7029365"] The best way of testing would just be to roll them a whole bunch of times and then do a statistical test on the plausibility of your results. Might take a while if you have a lot of dice. What I would do (and might do this weekend now I'm thinking about it - that's how you kill time in quarantine), is do it with a relatively low number of rolls for each, and then look in more detail at those which look suspicious. A chi-squared test is easy to do. Say you have a four-sided dice and roll out 40 times. You would expect each number to show up approximately 10 times. Count how many times each number appears. For each, subtract 10 (the expected value), square the result, and then divide that by ten. Add your four numbers together, and then find yourself a chi-squared table online (like [URL='https://people.richland.edu/james/lecture/m170/tbl-chi.html']here)[/URL] You would have 3 degrees of freedom (one less than the dice number), so find 3 on the leftmost column, then read across until you got whatever number you calculated. Looking at the top row at that point will tell you the probability of getting a result at least that far away from the expected value of your dice was fair. [/QUOTE]

Verification